In this renewal application for a Biomarker Development Laboratory, the investigators propose to use a carefully collected, curated and annotated bank of specimens to critically evaluate and compare a series of assays and lead markers to determine whether a clinically useful tool can be developed to augment mammography and ultrasound for the detection of breast cancer. The investigators believe early detection of breast cancer can only be improved by considering the problems inherent in the current best practice and supplying novel algorithms for solving these problems. Given the heterogeneity of the disease, its uncertain biologic progression, its high prevalence, and the variety of ways in which it is currently detected, our group considers it unlikely that any single approach will prove to be reliably effective in either improving upon or replacing current screening and diagnostic practices. Therefore, this proposal seeks to critically compare a series of molecular approaches that when applied in some combination will yield significant enhancement in the early detection of clinically relevant breast disease. The study will build upon and incorporate data from a separately funded NIH study that seeks to improve radiologic screening for breast cancer. Molecular markers will be incorporated with clinical and demographic data to build a comprehensive model for breast cancer prediction. This proposal combines the expertise and resources of Duke, Abbott Laboratories, Eastern Virginia Medical School, and University of Pittsburgh Medical Center.